HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
Hydrol. Earth Syst. Sci. Discuss., 10, 1–41, 2013www.hydrol-earth-syst-sci-discuss.net/10/1/2013/doi:10.5194/hessd-10-1-2013© Author(s) 2013. CC Attribution 3.0 License.
EGU Journal Logos (RGB)
Advances in Geosciences
Open A
ccess
Natural Hazards and Earth System
Sciences
Open A
ccess
Annales Geophysicae
Open A
ccess
Nonlinear Processes in Geophysics
Open A
ccess
Atmospheric Chemistry
and Physics
Open A
ccess
Atmospheric Chemistry
and Physics
Open A
ccess
Discussions
Atmospheric Measurement
Techniques
Open A
ccess
Atmospheric Measurement
Techniques
Open A
ccess
Discussions
Biogeosciences
Open A
ccess
Open A
ccess
BiogeosciencesDiscussions
Climate of the Past
Open A
ccess
Open A
ccess
Climate of the Past
Discussions
Earth System Dynamics
Open A
ccess
Open A
ccess
Earth System Dynamics
Discussions
GeoscientificInstrumentation
Methods andData Systems
Open A
ccess
GeoscientificInstrumentation
Methods andData Systems
Open A
ccess
Discussions
GeoscientificModel Development
Open A
ccess
Open A
ccess
GeoscientificModel Development
Discussions
Hydrology and Earth System
Sciences
Open A
ccess
Hydrology and Earth System
Sciences
Open A
ccess
Discussions
Ocean Science
Open A
ccess
Open A
ccess
Ocean ScienceDiscussions
Solid Earth
Open A
ccess
Open A
ccess
Solid EarthDiscussions
The Cryosphere
Open A
ccess
Open A
ccessThe Cryosphere
Discussions
Natural Hazards and Earth System
Sciences
Open A
ccess
Discussions
This discussion paper is/has been under review for the journal Hydrology and Earth SystemSciences (HESS). Please refer to the corresponding final paper in HESS if available.
Regional GRACE-based estimates ofwater mass variations over Australia:validation and interpretationL. Seoane1, G. Ramillien1, F. Frappart1, and M. Leblanc2
1Geosciences Environnement Toulouse GET – UMR5563, CNRS, Universite de ToulouseUPS, GRGS, 14, Avenue E. Belin, 31400 Toulouse, France2School of Earth & Environmental Sciences, James Cook University,Cairns, 4878, QLD, Australia
Received: 13 March 2013 – Accepted: 5 April 2013 – Published:
Correspondence to: L. Seoane ([email protected])
Published by Copernicus Publications on behalf of the European Geosciences Union.
1
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
Abstract
Time series of regional 2◦-by-2◦ GRACE solutions have been computed from 2003 to2011 with a 10 day resolution by using an energy integral method over Australia [112◦ E156◦ E; 44◦ S 10◦ S]. This approach uses the dynamical orbit analysis of GRACE Level 1measurements, and specially accurate along-track K Band Range Rate (KBRR) resid-5
uals (1 µm s−1 level of error) to estimate the total water mass over continental regions.The advantages of regional solutions are a significant reduction of GRACE aliasingerrors (i.e. north–south stripes) providing a more accurate estimation of water massbalance for hydrological applications. In this paper, the validation of these regionalsolutions over Australia is presented as well as their ability to describe water mass10
change as a reponse of climate forcings such as El Nino. Principal component anal-ysis of GRACE-derived total water storage maps show spatial and temporal patternsthat are consistent with independent datasets (e.g. rainfall, climate index and in-situobservations). Regional TWS show higher spatial correlations with in-situ water tablemeasurements over Murray–Darling drainage basin (80–90 %), and they offer a better15
localization of hydrological structures than classical GRACE global solutions (i.e. Level2 GRGS products and 400 km ICA solutions as a linear combination of GFZ, CSR andJPL GRACE solutions).
1 Introduction
The Gravity Recovery and Climate Experiment (GRACE) space gravity mission,20
launched in 2002, provides for the first time global estimates of changes in total waterstorage (surface water, soil moisture and groundwater) with unprecedented accuracyat global scales. GRACE data have already demostrated a strong potential for estim-ing hydrological information such as discharges (Syed et al., 2008), evapotranspiration(Rodell et al., 2004; Ramillien et al., 2006a), groundwater (Rodell et al., 2007; Leblanc25
et al., 2009) and extreme climate events as floods and droughts (Andersen et al., 2005;
2
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
Frappart et al., 2012, 2013; Houborg et al., 2012). Several analysis centres, includ-ing the Centre for Space Research, University of Texas at Austin (CSR), Jet Propul-sion Laboratory (JPL), GeoForschungsZentrum (GFZ) and Groupe de Recherche enGeodesie Spatiale (GRGS), use GRACE Level 1B observations to produce time seriesof global gravity solutions in terms of spherical harmonics. However, this type of repre-5
sentation of the Earth’s gravity field remains limited by errors with a north–south stripingpattern that degrades accuracy of the solutions, the effect of aliasing and the influenceof signals coming from regions outside the target basin. Filtering techniques can partlyremove some of these effects (Swenson et al., 2006; Klees et al., 2008; Frappart et al.,2011). New regional GRACE solutions using an energy integral method (Ramillien et10
al., 2011, 2012) yield to significant gains in the accuracy and resolution of total waterstorage estimates compared to those using the classical spherical harmonic approach.The present study aims to test these new regional solutions over the Australian con-tinent; the driest inhabited continent. In average the seasonal amplitude of TWS inAustralia is 3 times lower than that observed in the Amazon. The challenge is to detect15
realistic structures while noise amplitude in GRACE data is closer to the amplitude ofthe true water mass variations over the continent. Water storage variations from re-gional solutions over Australia are validated with independent information given by thein-situ observations (rainfall, groundwater levels). GRACE products have already beenused to study time variations of water mass storages and transfers over Australia. Syed20
et al. (2008) used GRACE data to estimate a depletion of 1.3 mm month−1 of total waterstorage in Australia. Leblanc et al. (2009) used GRACE global solutions to analyze theimpact of the Millennium drought on the water resources in the Murray–Darling Basin.Awange et al. (2009) noted the relatively small hydrological signal typical for much ofAustralia were not detectabled because of errors int the standard GRACE data pro-25
cessing (e.g. oceanic non-tidal model deficiencies) and filtering methods. Tregoning etal. (2008) showed the propagation of error of non-tidal oceanic modelling in the Gulfof Carpentaria onto the continent in northern Queensland and the Northern Territory.Frappart et al. (2011) proposed a filtering technique based on Independent Component
3
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
Analysis that better reproduced the in situ observations over the Murray–Darling Basinthan the spherical harmonic solutions obtained using other filtering.
Our regional approach using the energy integral method can potentially further im-prove the accuracy of GRACE TWS estimations. Over South America, Ramillien etal. (2012) have shown that thos regional approach offers a reduction of north–south5
striping due to mission configuration and temporal aliasing.The present study aims to test these new regional solution over the Australian con-
tinent. The challenge is to detect realistic structures while noise amplitude in GRACEdata is closer to the amplitude of the true water mass variations over the continent.
In the next section, all the datasets used are presented. Then, we discuss the results10
of the Principal Component Analaysis (PCA) separation in space and time modes ofGRACE TWS solutions in the third section. We interpret each of them in the terms of in-dividual hydrological structrures by the distinction of ground and surface water systemsand empirical climatic indexes. Analysis of GRACE signals over Australia using PCAhave already been proposed in previous studies (Reiser et al., 2010; Garcıa-Garcıa et15
al., 2011; Awange et al., 2011; Forootan and Kusche, 2012). For further validation, wecompared the new regional solutions to three in-situ and modelled hydrologic datasets:(1) rainfall across the continent; (2) surface and subsurface strorage in the Murray–Darling Basin; (3) streamflow data in the Fitzroy Basin (see Fig. 1). The conclusion ofour analysis is given in Sect. 4.20
2 Datasets
2.1 GRACE land waters products
2.1.1 Regional solutions
An alternative regional approach to the classical global one has been recently pro-posed to improve geographical localization of hydrological structures and to reduce25
4
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
leakage and aliasing. This energy integral method consists of recovering equivalent-water thickness of juxtaposed 2◦ or 4◦ geographical tiles by inversion of Differencesof Potential Anomaly (DPA) between GRACE vehicles A and B that are related to theGRACE inter-satellite KBR Range (KBRR) residuals (Ramillien et al., 2011, 2012).These KBRR residuals were obtained by correcting the raw observations from the a5
priori gravitational accelerations of known large-scale mass variations (i.e. atmosphereand oceanic mass variations, polar movements, as well as solid and oceanic tides)during the iterative least-squares orbit adjustment made by GINS software (Bruinsmaet al., 2010). The effects of non-conservative forces measured by on-board GRACEaccelerometers are also removed from the along-track observations, in order to ex-10
tract the contributions of not modelled phenomena, thus mainly water mass changes incontinental hydrology. Since classical gravimetric inversion does not provide a uniquesolution and to reduce the spurious effects of the noise in the observations, regulariza-tion strategies have been proposed to find numerically stable solutions, either basedon Singular Value Decomposition (SVD) and L curve analysis (Ramillien et al., 2011),15
or by introducing averaging radius as spatial constraints (Ramillien et al., 2012). Timeseries of successive 10 day regional solutions of water mass have been produced overa region included the whole Australia [112◦ E 156◦ E; 44◦ S 10◦ S]. In addition, for solv-ing the problem of the important long wavelength differences between DPA segmentsand stabilizing the inversion, low-degree polynomials are simply removed from each20
north–south DPA track to reduce long-term orbit error. Obviously, the risk of this op-eration is to loose useful long wavelength water mass signals that extensions exceedthe dimensions of the considered region. The numerical estimations show us that thepredicted regional solutions need to be completed by long wavelength components forcomparison with other data sets, when the geographical region is not large enough to25
contain these long gravity undulations (for more details see Ramillien et al., 2012).Preliminary tests were made to determine the harmonic degree of Level 2 GRGS
solutions to be used to complement short and medium wavelength regional solutions(Fig. 2). The degree of 6 explains mostly the RMS difference between global and
5
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
regional solutions. This optimal degree corresponds roughly to the latitudinal exten-sion of the region (≈6700 km). Before being added to the inverted GRACE signals,the long wavelength GRGS components represent about 47 and 45 % of the completehydrological signals over Australia and the Murray–Darling Basin in particular, respec-tively. As presented in Fig. 3 from December 2008 to November 2009, these seasonal5
amplitudes of the augmented 10 day regional solutions are typically of ±250 mm ofEWH in the tropical Northern region of Australia where the regular monsoon occurs inJanuary and April, yielding an alternation of wet and dry seasons. As expected, thesestrongest seasonal amplitudes are in the northern tropical band, for latitudes greaterthan −20◦ S. Moreover, the series of the regional TWS maps also reveals the constant10
loss of groundwater in the southeastern part of Australia which is affected by a droughtuntil 2010. During the dry period in the central desert (e.g. January to June 2006,where the cumulated rainfall over six consecutive months remains less than 60 mm),the amplitudes of our regional solutions remain lower than 30 mm of EWH (Fig. 4).As the expected values in this desert area should be small for this 6 month period,15
these recovered TWS amplitudes can be also considered as an indicator of the level ofuncertainty on our prediction.
2.1.2 Global Level 2 solutions from GRGS
To assess the differences of using regional approach instead of global solutions ortechniques we compare the solutions derived from integral energy approach to GRGS20
global solutions. The Level 2 GRGS-EIGEN-GL04-10-day models are derived fromLevel 1 GRACE measurements including KBRR and from LAGEOS-1/2 SLR data forenhancement of lower harmonic degrees (Bruinsma et al., 2010). These gravity fieldsare expressed in terms of normalized spherical harmonic coefficients from degree 2up to degree 50 using an empirical stabilization approach without any smoothing or25
filtering.In our study spatial means of Total Water Storage over an specific drainage basin
as Murray–Darling will be used in Sect. 3. We remember that for estimating the spatial6
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
means from the spherical harmonics approach, it is suitable to use the formula pro-posed earlier by Wahr et al. (1998) and Ramillien et al. (2006b). We computed the cor-responding variation of water volume inside the basin, δψ(∆t), as the scalar productof the water mass coefficients δCnm, δSnm with Anm and Bnm the normalized harmoniccoefficients of the considered geographical mask:5
δψ(∆t) = 4πR2N∑n=1
M∑m=0
(AnmδCnm(∆t) + BnmδSnm(∆t)) (1)
where R is the Earth’s mean radius (6371 km). TWS variations can be expressed interms of equivalent water height changes if δψ is divided by the area of the drainagebasin.
10 day Total Water Storage (TWS) grids of 1◦ spatial resolution are available for10
2002–2011 at http://grgs.obs-mip.fr.
2.1.3 Global Level 2 solutions from CSR, GFZ and JPL filtered using ICAapproach
A post-processing method based on Independent Component Analysis (ICA) was ap-plied to the Level 2 GRACE solutions from different official providers (i.e. UTCSR, JPL15
and GFZ) pre-filtered with 400 km radius Gaussian filters. This approach does not re-quire a priori information, except the assumption of statistical independence betweenthe elementary sources that compose the measured signals (i.e. useful geophysicalsignals plus noise). Separation consists of solving a linear system relating the GRACEsolutions provided for a given month, to the unknown independent sources. The con-20
tributors to the observed gravity field are forced to be uncorrelated numerically by im-posing diagonal cross-correlation matrices. Time series of ICA-based global maps ofcontinental water mass changes from combined UTCSR, JPL and GFZ GRACE solu-tions, computed over the period March 2003–December 2010, are used in this study.For a given month, the ICA-filtered solutions only differ from a scaling factor, so that25
7
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
the GFZ-derived ICA-filtered solutions are only presented. The efficiency of the ICAin separating gravity signals from noise by combining Level 2 GRACE solutions haspreviously been demonstrated over land (Frappart et al., 2010, 2011).
2.2 In-situ data and modelled hydrological datasets
2.2.1 BoM rainfall product5
To validate our regional solutions, we compared them to the rainfall over Australia. Weuse the monthly grids provide by the Bureau of Meteorology of the Australian Govern-ment (BoM). which provides rainfall grids expressed as Equivalent Water Height valuesfor the whole of Australia since 1967 (http://www.bom.gov.au). The rainfall product isbased on observations of irregularly distributed weather stations. They are averaged to10
monthly values and interpolated to geographic grids with the spatial resolution of 15′
by 15′ covering the region 111◦45′ E to 156◦30′ E in longitude and 9◦45′ S to 44◦45′ Sin latitude. The interpolation of the irregular data is based on the Barnes successive-correction method (Jeffrey et al., 2001).
2.2.2 Murray–Darling datasets15
Groundwater
In situ estimates of groundwater storage (GWS) in the Murray–Darling Basin wereobtained from an analysis of groundwater levels observed in government monitoringbores from 2002 to 2009. Assuming that (1) the shallow aquifers across the Murray–Darling drainage basin are hydraulically connected and that (2) at a large scale the frac-20
tured aquifers can be assimilated to a porous media, changes in groundwater storageacross the area can be estimated from observations of groundwater levels (e.g. Strass-berg et al., 2007). Variations in groundwater storage (∆SGW) were estimated from insitu measurements as:
8
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
∆SGW = Sy∆H (2)
where Sy is the aquifer specific yield (%) and H is the groundwater level (L−1) observedin monitoring bores. Groundwater level data (H) were sourced from Government de-partments of the States covered by the Murray Darling Basin (QLD, Natural Resourcesand Mines; NSW, Department of Water and Energy; VIC, Department of Sustainability5
and Environment; and SA, Department of Water Land and Biodiversity Conservation).Only observation bores (production bores excluded) with an average saturated zone<30 m from the bottom of the screened interval were selected as deeper bores canreflect processes occurring on longer time scales (Fetter, 2001). In the Murray andDarling drainage basins changes in groundwater levels were computed at three and10
six months time step respectively. Only bores with at least 80 % of the time periodspopulated with groundwater levels were selected for the analysis (1470 bores in theMurray and 958 bores in the Darling, Fig. 1). The median of the groundwater level foreach time period was first calculated at each bore and change at a bore was computedas the difference of median groundwater level between two consecutive periods. A spa-15
tial interpolation of the groundwater level change between two consecutive periods wasperformed across the basin using a kriging technique. Spatial averages of groundwa-ter level change were computed for each aquifer group. The Murray–Darling drainagebasin comprises several unconfined aquifers that can be regrouped into 3 groups ac-cording to their lithology. The specific yield Sy in Eq. (2) is estimated for each group: to20
range from 5 to 10%: for the clayey sand unconfined aquifer group (Macumber, 1999;Cresswell et al., 2003; Hekmeijer and Dawes, 2003a; CSIRO, 2008); from 10 to 15 %;for the shallow sandy clay unconfined aquifer group (Macumber, 1999; Urbano et al.,2004); and from 1 to 10 % for the fractured rock aquifer group (Cresswell et al., 2003;Hekmeijer and Dawes, 2003b; Smitt et al., 2003; Petheram et al., 2003). In-situ esti-25
mates of changes in GW storage are calculated using the spatially averaged changein groundwater level for aquifer group and the mean value of the specific yield for that
9
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
group. The range of possible values for the specific yield was used to estimate the un-certainty. Groundwater changes in the deep, confined aquifers (mostly Great ArtesianBasin and Renmark aquifers) are either due to: (1) a change in groundwater rechargeat unconfined outcrop; (2) shallow pumping at unconfined outcrop or (3) deep pumpingin confined areas for farming (irrigation and cattle industry). Total pumping from the5
deep, confined aquifers was estimated to amount to −0.42 km3 yr−1 in 2000 (Ife andSkelt, 2004) while groundwater pumping across the basin was −1.6 km3 in 2002–2003(Kirby et al., 2006). To allow direct comparison between TWS and in-situ GW estimates,pumping from the deep aquifers was added to the in-situ GW time series assuming the−0.42 km3 yr−1 pumping rate remained constant during the study period.10
Surface water
In the predominantly semiarid Murray–Darling Basin, most of the surface water (SW)stored in a network of reservoirs, lakes and weirs (Kirby et al., 2006). A daily timeseries of the surface water storage in the network of reservoirs, lakes, weirs and in-channel storage was obtained from the Murray–Darling Basin Commission and the15
state governments from 2002 to 2011.
Soil moisture from AWRA model
Soil moisture was derived from the Australian Water Resources Assessment (AWRA)hydrologic model (Van Dijk and Renzullo, 2010; Van Dijk et al., 2011). The systemcombines a comprehensive spatial hydrological model with meteorological forcing data20
and remotely sensed land surface properties.
2.2.3 Fitzroy streamflow dataset
A times series of monthly streamflow values spanning from 1998 to 2011 was ob-tained at two gauging stations along the Fitzroy River. The observations at Willare and
10
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
Noonbakhan stations (see Fig. 1) were provided by the departement of water (DoW) ofthe Governement of Western of Australia (www.water.wa.gov.au).
3 Results and discussion
3.1 PCA of the GRACE data sets
PCA was applied to the TWS time series of the 10 day GRACE-based regional, 10 day5
GRGS RL02 and monthly ICA 400 km filtered solutions, as well as the BoM rainfallgrids after having removed the dominant seasonal cycle using an averaging windowof 13 months. Figures 5, 7 and 9 present the spatial components of the three firstmodes that correspond to the most significant variability. The temporal componentsare presented in Figs. 6, 8 and 10. The three first mode of PCA explained a variance10
of 60, 13 and 15 % of variability respectively. Correlation values between the first modesof the different datasets are shown in Table 1 for the spatial patterns and Table 2 forthe temporal patterns and they will be discussed in the next sections.
3.1.1 The first mode of TWS variability
The first PCA mode explains 60, 67 and 52 % for the regional, GRGS and ICA 400 km15
filtered TWS solutions, respectively. The first spatial component of GRGS solutions stillcontains important north–south striping (Fig. 5) that degrades amplitudes and local-izations of the structures. Note that the regional approach reduces this polluting noisedue to orbit resonance by construction, since its first step consists of removing thelong-term satellite gravity undulations before inversion for equivalent-water heights, as20
explained in Sect. 2.1.1. For each PCA mode, the correlations between the spatialcomponents of GRACE base datasets and rainfall BoM product are shown in Table 1.Cross-correlations values for temporal components are also estimated and presentedin Table 2. For the first mode, temporal patterns correlations between different GRACEsolutions are in a good agreement (70–80 %). However, comparing GRACE data with25
11
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
rainfall product, only around 60 % of the signal are explained. In fact, GRACE satellitesobserve the mass variations due to rainfall but also other mass redistributions throughthe water balance equation. On the other hand, the correlation between the spatial pat-terns of GRACE products and rainfall drop due to the loss of mass only observed in thegravity field solutions at the Great Sandy Desert (at the northwest region of Australia)5
and the Murray Basin. In the Great Sandy desert, rainfall is nearly ungauged (Van Dijket al., 2011) and discrepancies are due to the missing information in the BoM products.The signal observed over Murray basin could be associated to aquifer variations. Theloss of mass over Murray–Darling Basin presented in the spatial component of the firstmode (Fig. 5) are due to the recent Millennium drought affecting this region (Leblanc10
et al., 2009, 2011). The corresponding negative trend on the temporal component canbe related to the increase of the drought from 2003 to 2009. The trend of water masschanges after 2010 indicates a gain of water mass in the southeastern part.
The South Oscillation Index (SOI) and the Pacific Decadal Oscillation (PDO) inter-annual variations were smoothed out using the same 13 month averaging window, and15
then compared to the temporal components of the first modes (Fig. 6a and b). As illus-trated by Table 3, high correlations (≈70 %) were found between these climatic indexesand the temporal components of the first TWS modes, suggesting the important influ-ence of these climate oscillations on the Australian water mass storage at interannualtime scales. Mantua and Hare (2002) compared PDO with rainfall data and suggested20
that warm phases of the PDO (positive values) coincide with anomalously dry periodsin eastern Australia. The signature of this phenomena can be seen in the spatial andtemporal components of the first mode of the PCA int the Murray–Darling basin (Figs. 5and 6) and it is probably the reason of the large negative correlations values obtainedin Table 3.25
The signifcant correlations obtained for the SOI and PDO suggest that the Australianclimat confirm that Australia climate is forced by these climate oscillations at interannualscales. In addition, water mass content reduction over central and southern regions of
12
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
Australia seems to be also associated to IOD positive phases as previously suggestedby Garcıa-Garcıa et al. (2010).
3.1.2 The second and third modes
In the case of our regional solutions, the second and third modes have very closeexplained variances (i.e. 13 and 15 %, respectively), so we decided to inverse their5
order for consistency with the modes from other solutions.The PCA spatial patterns of second mode of GRGS, ICA, rainfall data and third mode
of regional solutions presented in Fig. 7 show an important and continuous deficit ofmass occuring over the Murray–Darling and Lake Eyre regions. Looking at the temporalcomponents in Fig. 8a, we notice that the strongest deficit happened in April 2006 and10
it can be interpreted as the maxima of severity of the Millennium drought that has beenobserved in year 2006 (Leblanc et al., 2011).
However, the low correlation values (spatial or temporal, see Tables 1 and 2) be-tween GRACE products and rainfall over the whole Australia seem to indicate that themain amplitudes seen in the southeast part of Australia are due to other hydrological15
processes. In fact, a significant amplitude is located at Murray–Darling basin (Fig. 7spatial pattern and Fig. 8a temporal pattern) and probably linked to the interannualgroundwater level variations. For this purpose we derived groundwater internannualsignal from in situ observations for comparison (see Sect. 1.2). They are obtained bycomputing the average of observations over the Murray–Darling Basin, after the sea-20
sonal variations are removed using an averaging window. The water mass rate over theobservation period 2003–2010 are essentially presented in the first mode components,as in the previous section. Then, the linear trend is also removed from groundwater ob-servations in order to compare to the second mode temporal component. Signals areshown in Fig. 8a and b. Cross-correlation values are significant for the regional esti-25
mates (61 %) and GRGS solutions (73 %) with 4 months of time lag. In the case ofICA solutions, the correlation is lower and about 40 % with a 5 months of time lag. Thespatial PCA modes presented in Figs. 9 and 10a for temporal modes show positive
13
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
variations located in the North-West part of Australia which correspond to a gain ofwater mass related to an increase of rainfall (Li et al., 2009). As previously mentionedby Munier et al. (2012), an inter-annual cycle of TWS appears over the Fitzroy basinregion. This long-term undulation in the GRACE modes shown in Figs. 9 and 10a isconsistent with the third mode of rainfall (see correlation values in Tables 1 and 2).5
Cross-correlations values are 64, 50 and 74 % for regional, GRGS and ICA solutionsrespectively and with a time lag of 4 months with respect to rainfall data. Moreover,discharge records of Willare and Noonkanbah stations along the Fitzroy river havequasi-periodic peaks of water level in 2004, 2006 and 2008 (Fig. 10b), excepting for2010. These maxima of river discharge coincide quite well with the ones seen on the10
GRACE temporal mode, although the signature of the high Fitzroy river discharge of2009 is hardly present in GRACE data.
3.2 Localization of hydrological event
To demonstrate the improvement in detection brought by our regional solutions, weconfronted them with known meteorological events sharply localized in both space and15
time.
3.2.1 Exceptional rainfall in 2010
Anomaly maps of TWS were generated as the difference between the monthly TWSgrids corresponding to the exceptionally high rainfall in the central desert (from July toSeptember 2010), and the grid averaged over the previous dry season (i.e. mean of20
the six months from June to November 2009). GRACE-based TWS and BoM rainfallanomaly grids are presented in Fig. 11. When the BoM rainfall anomaly grids are takenas a reference, TWS anomaly grids from regional solutions offer a better spatial local-ization of the September 2010 rainfall structures than the anomalies computed withglobal GRGS solutions.25
14
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
3.2.2 Detection of the cyclone Charlotte
The heavy rainfall and subsequent floods that were caused by the cyclone Charlottebetween the 9 and 12 January 2009, in the Gulf of Carpentaria Basins, are clearlyvisible in Fig. 12. Tregoning et al. (2012) have already noticed the signature of this cli-mate event in GRACE global solutions. The first flood warnings for Queensland were5
issued in the first week of January from moderate to major flooding in the Gulf rivers,and Georgina and Diamantina rivers in Western Queensland (see Fig. 1). Since im-portant rainfall continued, these warnings were maintained for nine weeks (see BoMreport of 2009). Instantaneous TWS anomalies, defined as the difference of two con-secutive TWS 10 day regional and GRGS RL02 solutions are presented in Fig. 12 a10
and b respectively. In Fig. 12c TWS anomalies for ICA solutions defined as the differ-ence of two consecutive TWS monthly solutions. We notice clearly the strong signatureof the floods caused by the cyclone Charlotte, since the regional solutions reveal im-portant positive water heights of ≈600 mm of EWH, higher than the levels observed inthe previous months. For our regional solutions, this corresponding TWS anomaly is15
localized precisely in the Carpentaria drainage basins area (see Fig. 1). This suddenmeteorological event agrees well with the BoM rainfall data (Fig. 12d) and GRGS RL02estimates. However, ICA solutions show a very low amplitude of this event.
3.3 Basin-averaged time series of TWS
Leblanc et al. (2011) have shown the effects of drought over Murray–Darling Basin and20
they compared GRGS-RL02 global solutions to thus obtained as the sum of in situ datafor surface and ground waters and model outputs for soil moisture over the basin for athe period 2002–2010.
In our analysis in order to validate the regional approach, we compared the regionaland GRGS RL02 solutions and we included the ICA solutions. A comparison of time25
series of GRACE TWS averaged over the whole Murray–Darling using GRGS RL02,ICA and the regional solutions is presented in Fig. 13. We notice TWS variations
15
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
are consistent between all GRACE solutions (correlations values: ICA/regional=65 %,ICA/GRGS=70 %, regional/GRGS=90 %).
The average of rainfall in 2010, causde by a switch to a strong La Nina event, bringsthe end to this prolonged drought and leading the increase in TWS.
For comparison GRACE-based products with other independent datasets, we used5
the soil moisture outputs from AWRA Australian hydrological model and bores local-izations are extended then we have a better spatial coverage of groundwater levelvariations than in the case of the previous study of Leblanc et al. (2011). To maketheir spatial resolution compatible to the one of the regional GRACE solutions, TWSestimates reconstructed from in-situ and modelled hydrologic information were inter-10
polated onto 2◦ sampling grids. According to the temporal resolution of the GW datafor the period of 2003–mid 2009, the GRACE datasets were smoothed versus time:3 month and 6 month samplings for Murray and Darling basins, respectively. The cor-responding time series for the Murray and Darling Basins are presented in Fig. 14.Correlations of 0.7–0.9 and RMS values of 15–30 mm EWH from Table 4 suggest a15
good agreement between re-constructed TWS (i.e. SM+SW+GW) variations and theones from GRACE data sets, our regional GRACE solutions giving better results. Asseen in Fig. 14, we used a coefficient of proportionality (i.e. specific yield) rangingfrom 0.06 to 0.15 to convert in situ measurements of water table into water contentvariations, which are slightly larger but very comparable to the GRACE-based ampli-20
tudes. Gridded correlations and RMS differences in the Murray–Darling Basin are evenbetter for the regional GRACE solutions (Fig. 15), especially in the Murray Basin wherethe correlations are greater than 75 % and GW measurements are the densest (seeFig. 1); GRGS and ICA solutions may drop down to 25 % and even less.
16
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
4 Conclusions
In this paper, we validated our 10 day GRACE regional solutions of water mass vari-ations over Australia for the period 2003–2011 by comparing them with (i) globalGRGS RL02 and ICA 400 km solutions based on spherical harmonics representation,and (ii) independent datasets, i.e. estimates of rainfall, groundwater and surface wa-5
ter derived from in-situ observations and modelled soil moisture. Once applied to theGRACE data sets, PCA provided the main modes of variability of these solutions. Tem-poral and spatial patterns are consistent for the regional and global solutions. How-ever, the regional solutions offer better geographical localization of hydrological struc-tures, while global solutions remain polluted by aliasing errors (i.e. North–South strip-10
ing). High correlations between SOI and PDO climate indexes during the MillenniumDrought and our regional solutions exist, in particular for the first PCA mode. Secondand third modes are also related to the drought that occurred in the southeastern partof Australia up to 2006. While all the GRACE solutions remain consistent over Australia,our regional solutions yield the best agreement with in situ water table measurements15
in the Murray–Darling drainage Basin.
Acknowledgements. Lucia Seoane’s work was funded by ARC discovery projet at James CookUniversity (Australia). We would like to thank Adam Fakes for providing groundwater grids de-rived from in-situ measurements.
20
The publication of this article is financed by CNRS-INSU.
17
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
References
Andersen, O. B., Seneviratne, S. I., Hinderer, J., and Viterbo, P.: GRACE-derived terrestrialwater storage depletion associated with the 2003 European heat wave, Geophys. Res. Lett.,32, L18405, doi:10.1029/2005GL023574, 2005.
Awange, J., Sharifi, M., Baur, O., Keller, W., Featherstone, W., and Kuhn, M.: GRACE hydro-5
logical monitoring of Australia: Current limitations and future prospects, J. Spatial Sci., 54,23–36, doi:10.1080/14498596.2009.9635164, 2009.
Awange, J., Fleming, K. M., Kuhn, M., Featherstone, W. E., Heck, B., and Anjasmara, I.: Onthe suitability of the 4×4 GRACE mascon solutions for remote sensing Australian hydrology,Remote Sens. Environ., 115, 864–875, doi:10.1016/j.rse.2010.11.014, 2011.10
BoM report: Queensland Floods January and February 2009, Bureau of Meteorology, Queens-land, Australia, 2009.
Bruinsma, S., Lemoine, J.-M., Biancale, R., and Vales, N.: CNES/GRGS 10-daygravity models (release 2) and their evaluation, Adv. Space Res., 45, 587–601,doi:10.1016/j.asr.2009.10.012, 2010.15
Cresswell, R. G., Dawes, W. R., Summerell, G. K., and Walker, G. R.: Assessment of salinitymanagement options for Kyeamba Creek, New South Wales: Data analysis and groundwatermodelling, CSIRO Land and Water Technical Report 26/03, Murray Darling Basin Commis-sion, Canberra, 2003.
CSIRO Report: Water availability in the Loddon-Avoca, A report to the Australian Government20
from the CSIRO Murray-Darling Basin Sustainable Yields Project, Published by CSIRO, Aus-tralia, 123 pp., 2008.
Fetter, C. W.: Applied Hydrogeology, Published by Prentice Hall Inc, Upper Saddle River, NewJersey, 07458 USA, 2001.
Forootan, E. and Kusche, J.: Separation of global time-variable gravity signals into maximally25
independent components, J. Geodesy, 86, 477–497, doi:10.1007/s00190-011-0532-5, 2012.Frappart, F., Ramillien, G., Maisongrande, P., and Bonnet, M.-P.: Denoising satellite gravity
signals by independent component analysis, IEEE Geosci. Remote Sens. Lett., 7, 421–425,doi:10.1109/LGRS.2009.2037837, 2010.
Frappart, F., Ramillien, G., Leblanc, M., Tweed, S. O., Bonnet, M.-P., and Maisongrande, P.: An30
independent Component Analysis approach for filtering continental hydrology in the GRACEgravity data, Remote Sens. Environ., 115, 187–204, 2011.
18
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
Frappart, F., Papa, F., Santos da Silva, J., Ramillien, G., Prigent, C., Seyler, F., and Calmant, S.:Surface freshwater storage and dynamics in the Amazon basin during the 2005 exceptionaldrought, Environ. Res. Lett., 7, 044010, doi:10.1088/1748-9326/7/4/044010, 2012.
Frappart, F., Ramillien, G., and Ronchail, J.: Changes in terrestrial water storage versus rain-fall and discharges in the Amazon basin, Int. J. Climatol., online first, doi:10.1002/joc.3647,5
2013.Garcıa-Garcıa, D. C., Ummenhofer, C., Zlotnicki, V.: Australian water mass variations from
GRACE data linked to Indo-Pacific climate variability, Remote Sens. Environ., 115, 2175–2183, doi:10.1016/j.rse.2011.04.007, 2011.
Hekmeijer, P. and Dawes, W.: Assessment of salinity management options for South Loddon10
Plains, Victoria: Data analysis and groundwater modeling, CSIRO Land and Water TechnicalReport 24/03, Published by Murray-Darling Basin Commission, Canberra, Australia,, 2003a.
Hekmeijer, P. and Dawes, W.: Assessment of salinity management options for Axe Creek,Victoria: Data analysis and groundwater modelling, CSIRO Land and Water Technical Re-port 22/03, MDBC Publication 08/03, Published by Murray-Darling Basin Commission, Can-15
berra, Australia, 40 pp., 2003b.Houborg, R., Rodell, M., Li, B., Reichle, R., and Zaitchik, B. F.: Drought indicators based on
model-assimilated Gravity Recovery and Climate Experiment (GRACE) terrestrial water stor-age observations, Water Resour. Res., 48, W07525, doi:10.1029/2011WR011291, 2012.
Ife, D. and Skelt, K.: Murray-Darling Basin Groundwater Status, Murray-Darling Basin Commis-20
sion, publication 32/04, Canberra, ISBN 1876830948, 2004.Jeffrey, S. J., Carter, J. O., Moodie, K. B., and Beswick, A. R.: Using spatial interpolation to
construct a comprehensive archive of Australian climate data, Environ. Model. Softw., 16,309–330, 2001.
Klees, R., Revtova, E. A., Gunter, B. C., Ditmar, P., Oudman, E., Winsemius, H. C., and25
Savenije, H. H. G.: The design of an optimal filter for monthly GRACE gravity models, Geo-phys. J. Int., 175, 417–432, doi:10.1111/j.1365-246X.2008.03922.x, 2008.
Kirby, M., Evans, R., Walker, G., Cresswell, R., Coram, J., Khan, S., Paydar, Z., Mainuddin,M., McKenzie, N., and Ryan, S.: The shared water resources of the Murray–Darling Basin,Publication 21/06, available online at: http://www.mdbc.gov.au, Murray–Darling Basin Com-30
mission, Canberra, 2006.
19
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
Leblanc, M., Tregoning, P., Ramillien, G., Tweed, S., and Fake, A.: Basin scale, integrated ob-servations of the early 21st Century multi-year drought in southeast Australia, Water Resour.Res., 45, W04408, doi:10.1029/2008WR007333, 2009.
Leblanc, M., Tweed, S., Ramillien, G., Tregoning, P., Frappart, F., Fakes, A., and Cartwright, I.:Groundwater change in the Murray basin from long-term in-situ monitoring and GRACE es-5
timates, in: Climate change effects on groundwater resources: A global synthesis of findingsand recommendations, edited by: Treidel, H. and Gurdak, J. J., CRC Press, 169–187, 2011.
Li, L. T., Donohue, R. J., McVicar, T. R., Van Niel, T. G., Ten, J., Potter, N. J., Smith, I. N.,Kirono, D. G. C., Bathols, J. M., Cai, W., Marvanek, S. P., Gallant, S. N., Chiew, F. H. S.,and Frost, A. J.: Climate data and their characterisation for hydrological scenario modelling10
across northern Australia. A report to the Australian Government from the CSIRO North-ern Australia Sustainable Yields Project. Published by CSIRO Water for a Healthy CountryFlagship, Australia, 63 pp., 2009.
Macumber, P. G.: Groundwater flow and resource potential in the Bridgewater and SalisburyWest GMAs, Phillip Macumber Consulting Services, Melbourne, 88 pp., 1999.15
Mantua, N. J. and Hare, S. R.: The Pacific Decadal Oscillation, J. Oceanogr., 58, 35–44, 2002.Munier, S., Becker, M., Maisongrande, P., and Cazenave, A.: Using GRACE to detect Ground-
water Storage variations: the cases of Canning Basin and Guarani aquifer system, Int. WaterTechnol. J., Vol. 2, 2012.
Petheram, C., Dawes, W., Walker, G., and Grayson, R. B.: Testing in class variabil-20
ity of groundwater systems: Local upland systems, Hydrol. Process., 17, 2297–2313,doi:10.1002/hyp.1333, 2003.
Ramillien, G., Frappart, F., Guntner, A., Ngo-Duc, T., Cazenave, A., and Laval, K.:Time variations of the regional evapotranspiration rate from Gravity Recovery andClimate Experiment (GRACE) satellite gravimetry, Water Resour. Res., 42, W10403,25
doi:10.1029/2005WR004331, 2006a.Ramillien, G., Lombard, A., Cazenave, A., Ivins, E. R., Llubes, M., Remy, F., and Biancale, R.:
Interannual variations of the mass balance of the Antarctica and Greenland ice sheets fromGRACE, Global Planet. Change, 53, 198–208, doi:10.1016/j.gloplacha.2006.06.003, 2006b.
Ramillien, G., Biancale, R., Gratton, S., Vasseur, X., and Bourgogne, S.: GRACE-derived sur-30
face mass anomalies by energy integral approach, Application to continental hydrology, J.Geodesy., 85, 313–328, doi:10.1007/s00190-010-0438-7, 2011.
20
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
Ramillien, G., Seoane, L., Frappart, F., Biancale, R., Gratton, S., Vasseur, X., and Bour-gogne, S.: Constrained regional recovery of continental water mass time-variations fromGRACE based geopotential anomalies over South America, Surv. Geophys., 33, 887–905doi:10.1007/s10712-012-9177-z, 2012.
Rieser, D., Kuhn, M., Pail, R., Anjasmara, I. M., and Awange, J.: Relation between GRACE-5
derived surface mass variations and precipitation over Australia, Aust. J. Earth Sci., 57, 887–900, doi:10.1080/08120099.2010.512645, 2010.
Rodell, M., Famiglietti, J. S., Chen, J., Seneviratne, S. I., Viterbo, P., Holl, S., and Wilson, C. R.:Basin scale estimate of evapotranspiration using GRACE and other observations, Geophys.Res. Lett., 31, L20504, doi:10.1029/2004GL020873, 2004.10
Rodell, M., Chen, J., Kato, H., Famiglietti, J. S., Nigro, J., and Wilson, C.: Estimating groundwa-ter storage changes in the Mississippi River basin (USA) using GRACE, Hydrogeol. J., 15,159–166, doi:10.1007/s10040-006-0103-7, 2007.
Smitt, C., Doherty, J., Dawes, W., and Walker, G.: Assessment of salinity management optionsfor the Brymaroo catchment, South-eastern Queensland, CSIRO Land and Water Technical15
Report 23/03, Murray-Darling Basin Commission, Canberra, Australia, 2003.Strassberg, G., Scanlon, B. R., and Rodell, M.: Comparison of seasonal terrestrial water
storage variations from GRACE with groundwater-level measurements fromthe High PlainsAquifer (USA), Geophys. Res. Lett., 34, L14402, doi:10.1029/2007GL030139, 2007.
Swenson, S. C. and Wahr, J.: Post-processing removal of correlated errors in GRACE data,20
Geophys. Res. Lett., 33, L08402, doi:10.1029/2005GL025285, 2006.Syed, T. H., Famiglietti, J. S., Rodell, M., Chen, J., and Wilson, C. R.: Analysis of terres-
trial water storage changes from GRACE and GLDAS, Water Resour. Res., 44, W02433,doi:10.1029/2006WR005779, 2008.
Tregoning, P., Lambeck, K., and Ramillien, G.: GRACE estimates of sea surface height25
anomalies in the Gulf of Carpentaria, Australia, Earth Planet. Sc. Lett., 271, 241–244,doi:10.1016/j.epsl.2008.04.018, 2008.
Tregoning, P., McClusky, S., van Dijk, A. I. J. M., Crosbie, R. S., and Pena-Arancibia, J. L.:Assessment of GRACE Satellites for Groundwater Estimation in Australia, National WaterCommission, Canberra, 82 pp., 2012.30
Urbano, L. D., Person, M., Kelts, K., and Hanor, J. S. :Transient groundwater impacts on thedevelopment of paleoclimatic lake records in semi-arid environments, Geofluids, 4, 187–196,doi:10.1111/j.1468-8123.2004.00081.x, 2004.
21
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
van Dijk, A. I. J. M. and Renzullo, L. J.: Water resource monitoring systems and the role of satel-lite observations, Hydrol. Earth Syst. Sci., 15, 39–55, doi:10.5194/hess-15-39-2011, 2011.
van Dijk, A. I. J. M., Renzullo, L. J., and Rodell, M.: Use of GRACE terrestrial water storageretrievals to evaluate model estimates by the Australian water resources assessment system,Water Resour. Res., 47, W11524, doi:10.1029/2011WR010714, 2011.5
Wahr, J., Molenaar, M., and Bryan, F.: Time variability of the Earth’s gravity field: hydrologi-cal and oceanic effects and their possible detection using GRACE, J. Geophys. Res., 103,30205–30229, 1998.
22
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
Table 1. Correlation values between PCA spatial components (mode 1, 2 and 3) of differentsGRACE based datasets and rainfall BoM product.
Mode 1 Mode 2a Mode 3b
Regional solution/GRGS 0.78 0.93 0.88Regional solution/ICA 0.86 0.87 0.86GRGS/ICA 0.82 0.83 0.77Regional solution/rainfall 0.44 0.38 0.50GRGS/rainfall 0.38 0.24 0.55ICA/rainfall 0.67 0.62 0.55
a Except for regional solutions: here mode 3 is used to estimate correlationvalues. b Except for regional solutions: here mode 2 is used to estimatecorrelation values.
23
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
Table 2. Cross-correlation maxima values and corresponding temporal lag in number of months(second value in italic) between PCA temporal components (mode 1, 2 and 3) of differentsGRACE based datasets regional solutions, GRGS and ICA as well as rainfall BoM product.Data derived from Regional and GRGS are filtered and interpolated to monthly resolution. Thetime lag represents the time delay between water mass variations derived from GRACE andthe observed rainfall.
Mode 1 Mode 2a Mode 3b
Regional solution/GRGS 0.98, 0 0.96, 0 0.94, 0Regional solution/ICA 0.95, 0 0.93, 0 0.79, 0GRGS/ICA 0.97, 0 0.89, 0 0.64, 0Regional solution/rainfall 0.59, −1 0.19, 6 0.64, 4GRGS/rainfall 0.64, −1 0.24, 5 0.50, 4ICA/rainfall 0.70 , 0 0.33, 1 0.73, 4
a Except for regional solutions: here mode 3 is used to estimate correlationvalues. b Except for regional solutions: here mode 2 is used to estimatecorrelation values.
24
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
Table 3. Correlation values between climate index and PCA temporal patterns for modes 1, 2and 3 of regional, GRGS and ICA solutions.
SOI Index PDO index
Mode 1
Regional 0.73 −0.80GRGS 0.70 −0.78ICA 0.74 −0.72
25
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
Table 4. Correlation (first value) and RMS of the difference (second value in italic) between timeseries over Murray and Darling computed from different GRACE solutions and the modeled/in-situ estimates (SM+SW+GW) using the maximum, mean and minimum values of the specificyield. RMS units are mm of EWH.
Murray
Regional GRGS ICA
Symin0.92, 16.27 0.87, 20.27 0.87, 24.96
Symean0.92, 24.03 0.88, 25.10 0.85, 36.03
Symax0.92, 34.35 0.89, 33.72 0.84, 47.47
Darling
Regional GRGS ICA
Symin0.71, 12.18 0.68, 13.40 0.53, 14.34
Symean0.70, 13.17 0.67, 14.20 0.48, 16.30
Symax0.68, 14.63 0.65, 15.53 0.43, 18.44
26
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
120˚
120˚
130˚
130˚
140˚
140˚
150˚
150˚
−40˚ −40˚
−30˚ −30˚
−20˚ −20˚
−10˚ −10˚
(1) (2)
Fitzroy river
Murray−Darlingbasin
Gulf of Carpentariadrainage area
Georgina river
Diamantina river
Fig. 1. Australian drainage regions used in our study. At Fitzroy river, two stations (discharge)are considered: (1) Willare station (2) Noonkanbah station. Geographical locations of the boresin the Murray–Darling are marked as black cercles. They represented, with at least 60 % of dataperiod populated (because of their availability during the period 2002–2009), averages of thefirst 30 m of the water column height.
27
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
Fig. 2. Water mass time series averaged over Australia: (a) difference between regional andGRGS solutions, (b) GRGS GRACE solutions up to degree 6, and (c) their difference (RMSvalue 39 mm).
28
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
−40˚
−20˚
−40˚
−20˚
120˚ 140˚
−40˚
−20˚
120˚ 140˚ 120˚ 140˚ 120˚ 140˚ 120˚ 140˚ 120˚ 140˚−210
−180
−150
−120
−90
−60
−30
0
30
60
90
120
150
180
210
mm
EW
H
04/12/08−13/12/08 14/12/08−23/12/08 24/12/08−02/01/09 03/01/09−12/01/09 13/01/09−22/01/09 23/01/09−01/02/09
02/02/09−11/02/09 12/02/09−21/02/09 22/02/09−03/03/09 04/03/09−13/03/09 14/03/09−23/03/09 24/03/09−02/04/09
03/04/09−12/04/09 13/04/09−22/04/09 23/04/09−02/05/09 03/05/09−12/05/09 13/05/09−22/05/09 23/05/09−01/06/09
−40˚
−20˚
−40˚
−20˚
120˚ 140˚
−40˚
−20˚
120˚ 140˚ 120˚ 140˚ 120˚ 140˚ 120˚ 140˚ 120˚ 140˚−210
−180
−150
−120
−90
−60
−30
0
30
60
90
120
150
180
210
mm
EW
H
02/06/09−11/06/09 12/06/09−21/06/09 22/06/09−01/07/09 02/07/09−11/07/09 12/07/09−21/07/09 22/07/09−31/07/09
01/08/09−10/08/09 11/08/09−20/08/09 21/08/09−30/08/09 31/08/09−09/09/09 10/09/09−19/09/09 20/09/09−29/09/09
30/09/09−09/10/09 10/10/09−19/10/09 20/10/09−29/10/09 30/10/09−08/11/09 09/11/09−18/11/09 19/11/09−28/11/09
Fig. 3. Time series of regional maps of water mass from December 2008 to November 2009.We can notice the annual signal with amplitude of ±250 mm of EWH representing the alternat-ing wet and dry seasons.
29
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
2005.5 2005.8 2006 2006.2 2006.5 2006.8 2007-30
-20
-10
0
10
20
30
40
50
mm
EW
H
-30
-20
-10
0
10
20
30
40
50
mm
of
mon
thly
rai
nfal
l
Rainfall histogramRegional Solution (sigma=12.6 mm)ICA filtered solution (sigma=6.9 mm)GRGS RL02 (sigma=10.3 mm)
TWS mean over a dry region in the period 01/2006-06/2006
Fig. 4. Time series of TWS means over the estimated desert region (Central Australia) of18 tiles, the largest signal amplitudes are lesser than 30 mm and the RMS value is of 12.24 mm.
30
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
120˚ 130˚ 140˚ 150˚
−40˚ −40˚
−30˚ −30˚
−20˚ −20˚
−50
−40
−30
−20
−10
0
10
20
30
40
50
mm EWH
−50
−40
−30
−20
−10
0
10
20
30
40
50
mm EWH
120˚ 130˚ 140˚ 150˚
−40˚ −40˚
−30˚ −30˚
−20˚ −20˚
120˚ 130˚ 140˚ 150˚
−40˚ −40˚
−30˚ −30˚
−20˚ −20˚
−50
−40
−30
−20
−10
0
10
20
30
40
50
mm EWH
−50
−40
−30
−20
−10
0
10
20
30
40
50
mm
120˚ 130˚ 140˚ 150˚
−40˚ −40˚
−30˚ −30˚
−20˚ −20˚
Mode 1 TWS Regional Mode 1 TWS GRGS RL02
Mode 1 TWS GFZ ICA Mode 1 Rainfall
Fig. 5. Spatial components of PCA 1st orthogonal mode.
31
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
2003 2004 2005 2006 2007 2008 2009 2010 2011-1
-0.5
0
0.5
1 Regional solution Mode 1 (Var=0.60)GRGS RL02 Mode 1 (Var= 0.67)GFZ ICA Mode 1 (Var=0 .52)BoM Rainfall Mode 1 (Var= 0.33)
(a) PCA
2003 2004 2005 2006 2007 2008 2009 2010 2011-15
-10
-5
0
5
10
15SOI interannualPDO interannual (scale x 10)
(b) Climate Index
Fig. 6. (a) Temporal components of PCA 1st orthogonal mode. (b) Interannual variations of theSouthern Oscillation Index (SOI) and Pacific Decadal Oscillation (PDO).
32
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
120˚ 130˚ 140˚ 150˚
−40˚ −40˚
−30˚ −30˚
−20˚ −20˚
−50
−40
−30
−20
−10
0
10
20
30
40
50
mm EWH
−50
−40
−30
−20
−10
0
10
20
30
40
50
mm EWH
120˚ 130˚ 140˚ 150˚
−40˚ −40˚
−30˚ −30˚
−20˚ −20˚
120˚ 130˚ 140˚ 150˚
−40˚ −40˚
−30˚ −30˚
−20˚ −20˚
−50
−40
−30
−20
−10
0
10
20
30
40
50
mm EWH
−50
−40
−30
−20
−10
0
10
20
30
40
50
mm
120˚ 130˚ 140˚ 150˚
−40˚ −40˚
−30˚ −30˚
−20˚ −20˚
Mode 3 TWS Regional Mode 2 TWS GRGS RL02
Mode 2 TWS GFZ ICA Mode 2 Rainfall
Fig. 7. Spatial components of PCA 3rd orthogonal mode of regional solutions and 2nd modeof GRGS, ICA and Rainfall product.
33
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
2003 2004 2005 2006 2007 2008 2009 2010 2011-1
-0.5
0
0.5
1
Regional solution Mode 3 (Var=0.13)GRGS RL02 Mode 2 (Var= 0.17)GFZ ICA Mode 2 (Var= 0.18)BoM Rainfall Mode 2 (Var= 0.27)
(a) PCA
2003 2004 2005 2006 2007 2008 2009 2010 2011-0.2
-0.1
0
0.1
0.2
met
res
Murray- Darling BasinMurray BasinDarling Basin
(b) Groundwater level observations over Murray-Darling
Fig. 8. (a) Temporal components of PCA 3rd orthogonal mode of regional solutions and2nd mode of GRGS, ICA and Rainfall product. (b) Groundwater levels from in-situ measure-ments over Darling basin.
34
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
120˚ 130˚ 140˚ 150˚
−40˚ −40˚
−30˚ −30˚
−20˚ −20˚
−50
−40
−30
−20
−10
0
10
20
30
40
50
mm EWH
−50
−40
−30
−20
−10
0
10
20
30
40
50
mm EWH
120˚ 130˚ 140˚ 150˚
−40˚ −40˚
−30˚ −30˚
−20˚ −20˚
120˚ 130˚ 140˚ 150˚
−40˚ −40˚
−30˚ −30˚
−20˚ −20˚
−50
−40
−30
−20
−10
0
10
20
30
40
50
mm EWH
−50
−40
−30
−20
−10
0
10
20
30
40
50
mm
120˚ 130˚ 140˚ 150˚
−40˚ −40˚
−30˚ −30˚
−20˚ −20˚
Mode 2 TWS Regional Mode 3 TWS GRGS RL02
Mode 3 TWS GFZ ICA Mode 3 Rainfall
Fig. 9. Spatial components of PCA 2nd orthogonal mode of regional solutions and 3rd modeof GRGS, ICA and Rainfall product.
35
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
2003 2004 2005 2006 2007 2008 2009 2010 2011-1
-0.5
0
0.5
1
Regional solution Mode 2 (Var= 0.15)GRGS RL02 Mode 3 (Var= 0.09)GFZ ICA Mode 3 (Var= 0.10)BoM Rainfall Mode 3 (Var= 0.13)
(a) PCA
2003 2004 2005 2006 2007 2008 2009 2010 2011Years
0
1e+06
2e+06
3e+06
4e+06
5e+06
6e+06
7e+06
8e+06
9e+06
1e+07
Tot
al S
trea
m D
isch
arge
Vol
ume
in M
egal
itres
Willare StationNoonkanbah station
(b) River DischargeFitzroy River
Fig. 10. (a) Temporal components of PCA 2nd orthogonal mode of regional solutions and3rd mode of GRGS, ICA and Rainfall product. (b) In-situ observations of Fitzroy river dischargein MegaLitres (stations Willare and Noonkanbah).
36
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
120˚ 135˚ 150˚
−30˚
−15˚
120˚ 135˚ 150˚
−30˚
−15˚
120˚ 135˚ 150˚
−30˚
−15˚
120˚ 135˚ 150˚
−30˚
−15˚
120˚ 135˚ 150˚
−30˚
−15˚
120˚ 135˚ 150˚
−30˚
−15˚
120˚ 135˚ 150˚
−30˚
−15˚
−150
−100
−50
0
50
100
150
mm
EW
H
120˚ 135˚ 150˚
−30˚
−15˚
0
50
100
150
200
mm
09−201007−2010(a) Regional solution
(b) GRGS solution
(c) ICA solution
(d) Rainfall
Fig. 11. Regional and GRGS RL02 of ∆TWS monthly maps estimated with respect to the meanmap for the previous driest season (from June to November 2009), and rainfall maps. For thesetwo periods we can notice that the regional solutions show a better localization of better theexceptional rainfall maximum ocurring in the central region of Australia.
37
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
130˚ 135˚ 140˚ 145˚ 150˚
−20˚
−15˚
−10˚
−600
−400
−200
0
200
400
600
mm EWH/month
(a) Delta TWS from regional approach estimated as :
TWS ( 24−12−08 / 02−01−09 ) minus TWS ( 03−01−09 / 12−01−09 )
130˚ 135˚ 140˚ 145˚ 150˚
−20˚
−15˚
−10˚
−600
−400
−200
0
200
400
600
mm EWH/month
(b) Delta TWS from GRGS solution estimated as :
TWS ( 24−12−08 / 02−01−09 ) minus TWS ( 03−01−09 / 12−01−09 )
130˚ 135˚ 140˚ 145˚ 150˚
−20˚
−15˚
−10˚
−600
−400
−200
0
200
400
600
mm/month
(c) Delta TWS derived from ICA (January 2009 minus December 2008)
130˚ 135˚ 140˚ 145˚ 150˚
−20˚
−15˚
−10˚
−600
−400
−200
0
200
400
600
mm/month
(d) BoM Rainfall anomaly (January 2009 minus December 2008)
Fig. 12. ∆TWS time series map derived from regional solutions (a), from GRGS RL02 (b),from ICA solutions (c) and rainfall anomalies for January 2009 (d). We can notice on Januarysolutions the strong amplitudes in the Gulf of Carpentaria due to the floods caused by thecyclone Charlotte (8 January–12 January 2009).
38
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
Fig. 14. 10-days and monthly time series of averaged total water storage over thewhole Murray-Darling Basin
from 2003 to 2011 computed from 10-day GRGS global solutions (red line), 10-day regional solutions (blue
line) and monthly ICA 400-km filtered solutions (green line).
33
Fig. 13. 10 days and monthly time series of averaged total water storage over the wholeMurray–Darling Basin from 2003 to 2011 computed from 10 day GRGS global solutions (redline), 10 day regional solutions (blue line) and monthly ICA 400 km filtered solutions (greenline).
39
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
Fig. 15. (a) 3-month averaged time series of TWS over Murray Basin and (b) 6-month averaged time series
of TWS over Darling basin for 2003-2010. Series are computed from GRACE-based regional solutions (blue
line), GRGS RL02 (red line), ICA 400-km filtered solutions (green line) and TWS estimated as groundwater
storage + surface water observations + modeled soil moisture (grey lines). Discountionous grey line: Maximum
and minimum values of the specific yield are used (values of 0.06 and 0.14); Continiuous grey line: mean value
of specific yield is used (0.01).
34
Fig. 14. (a) 3 month averaged time series of TWS over Murray Basin and (b) 6 month averagedtime series of TWS over Darling basin for 2003–2010. Series are computed from GRACE-based regional solutions (blue line), GRGS RL02 (red line), ICA 400 km filtered solutions (greenline) and TWS estimated as groundwater storage+ surface water observations + modeled soilmoisture (grey lines). Discountionous grey line: Maximum and minimum values of the specificyield are used (values of 0.06 and 0.14); Continiuous grey line: mean value of specific yield isused (0.01).
40
HESSD10, 1–41, 2013
RegionalGRACE-derived watermass variations over
Australia
L. Seoane et al.
Title Page
Abstract Introduction
Conclusions References
Tables Figures
J I
J I
Back Close
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion
Paper
|D
iscussionP
aper|
Discussion
Paper
|D
iscussionP
aper|
CORRELATION
140˚ 150˚−40˚
−30˚
−20˚
RMS
140˚ 150˚
140˚ 150˚−40˚
−30˚
−20˚
140˚ 150˚
140˚ 150˚−40˚
−30˚
−20˚
0.00 0.25 0.50 0.75 1.00
140˚ 150˚
0 10 20 30 40 50 60 70 80
mm
(a) Regional solution
(b) GRGS RL02 solution
(c) GFZ ICA solution
Fig. 15. Spatial correlation (left maps) and RMS (right maps) of differences between TWS (in-situ GW measurements+ surface water+modeled soil moisture) and: (a) regional solutions,(b) GRGS RL02 solutions and (c) ICA 400 km filtered solutions.
41